Stroke induces a cascade of neurophysiologic changes in cortical and spinal circuits that result in biomechanical gait impairments (reduced paretic propulsion, footdrop) and gait dysfunction (reduced speed). While increasing gait speed is a major goal of stroke gait rehabilitation, targeting walking speed as a primary gait rehabilitation outcome without regard to biomechanical and neural mechanisms fails to meet the emerging standards of precision medicine, which is the future of rehabilitation research. Thus, here, we will confirm a novel theoretical framework regarding neurobiological (top-down) and biomechanics (bottom-up) mechanisms of how 2 gait treatments improve walking speed post-stroke. Fast treadmill walking (Fast), a well- studied and clinically-used intervention, improves gait speed. However, Fast-induced speed improvements in people post-stroke may occur at the cost of inter-limb asymmetry, energy inefficiency, and maladaptive neuroplasticity. Recent work has demonstrated that combining Fast with functional electrical stimulation (FastFES) not only leads to improvements in gait speed but also reduces energy cost (EC) of stroke gait. Because reduced EC is crucial for sustaining faster gait speeds and promoting community activity, biomechanical factors influencing EC post-stroke merit more in-depth study. Building upon knowledge gained from previous FastFES work and our preliminary data, Aim 1 will test our hypothesis that in contrast to Fast, FastFES promotes greater use of the paretic leg for forward propulsion, thereby improving inter-limb biomechanical asymmetry, which we hypothesize reduces EC. Gait rehabilitation essentially involves retraining the central nervous system. Our lack of understanding of neuroplasticity mechanisms underlying gait interventions continues to be a barrier to improving gait rehabilitation outcomes. Aim 2 will determine, for the first time, if and how FastFES and Fast modulate excitability of neural circuits impacted by stroke and implicated in locomotor control. Stroke leads to decrease in lesioned motor cortex (M1) excitability and corticospinal tract (CST) output, and elevated spinal reflex excitability. New findings from our lab suggest that unlike Fast, FastFES enhances lesioned CST and M1 excitability, restoring more normal CST output. FastFES and Fast also differ in their effects on spinal excitability. Like most gait treatments, Fast and FastFES must contend with high inter-individual variability in treatment responses (a subset of participants are ?non- responders?). Aim 3 will address whether baseline measures or short-term changes in neurophysiological biomarkers (CST and spinal excitability) can predict long-term training-effects. Results from our mechanism- focused clinical investigation will elucidate how, why, and for whom Fast and FastFES induce clinical benefits. The overall impact of this work will be future development of cutting-edge gait treatments that are individually- tailored based on neurobiological, biomechanical, and clinical characteristics to improve both gait quality and gait function, well-aligned with the NICHD mission of improving health through ?optimal? rehabilitation.